@adapalmer.bsky.social a question for you about history, GenAI, and public trust. It seems to me that most of the fear I see about how GenAI will erode public trust are making arguments from an assumption that we've always had an ability to trust media evidence (photo/video/audio) as factual.

Oregon

Good morning. Immigration is good. Trade is good. There are nuances but they are narrow enough that free and open movement of people and goods should be the default.

@dys-morphia.bsky.social plenty of the west coast has lots of water! You just have to get up to the Cascade range. (yes, I also don't think about water much, except during the annual July/Aug dry spell. The first big rain of September is a release of pent-up stress you forgot you were carrying).

@papamurphyspizza.bsky.social Can you help me understand why your app is by far the most paranoid on my phone? I feel like there must be an incredible story behind the absolute refusal to let android developers order take 'n bake pizza.

I've been thinking about this a lot, and am hung up on a particular point. Namely, that for all the talk about mechanizing thought, it seems that most of the benefits of computers has come from machines "doing things that humans are bad it". And we haven't made much progress at recreating thought!

As a self-professed methane pipe enthusiast, I want to take a moment to defend: 1) Natural gas as a service and 2) Natural gas for cooking in particular.

I want to stress that while playing a game of "find the most mundane task that stumps the billion dollar LLM" is very *entertaining* it doesn't really demonstrate anything of value. The true measure of any tool is simply: "Is it useful?" All other metrics hang on this question.

I feel like I've been given some super powered version of ChatGPT that is immune to all the problems everyone else has with the tools. 🤷‍♂️ chatgpt.com/share/689601...

🤷‍♂️ chatgpt.com/share/689579...

How is it acceptable that a journalist can write 500 words about the valuation and sale of a website and NOT ONCE link to the site under discussion?? How is this even remotely acceptable as a journalistic practice? (To be clear, this happens all the time, I use this example because I saw it today).

I checked in with my wife (romance novel enthusiast) about this and have learned that most women find bare chested cover art to be distasteful. Or any cover art that depicts the characters. This is because it interferes with the reader's ability to imagine for themselves their own characters.

School supplies are baffling to me. Our district furnishes every child with an Ipad or chromebook as part of enrollment. AND YET, every year I'm asked to go out and supply, from my own pocket, glue sticks for the class to use.

Can all the people talking about female gaze explain romance novel covers to me? I understand they've been trending away from this style of cover art (you should see my wife's collection of 90's/00's novels!), but still! Square this circle for me.

In general, I've found LLMs to be *amazing* at helping with what I call "problems that have already been solved". That is, a problem that has already been encounted by many other people and a solution found and documented. It's for this reason that it's especially good at helping with homework.

It's kinda amazing to me that so many people still hold the stance of "how could LLMs possibly be useful for anything??" I continue to find very useful applications for them in my day to day life. The next time you're trying to fix some bizarre household problem try feeding the symptoms into an LLM

And the logic is stupidly simple. Let people move to places where they can be more productive. That's it. Imagine we disallowed this inside the country. It'd be a disaster. Massive wealth destruction. But we do it every day worldwide.

Science fiction writers all have this fascination with the word "creche" and I have no idea why. Example: questionablecontent.net/view.php?com...

In other words, I don't think AI is the problem. Students were already doing things like taking the easiest possible classes, using chegg, begging teachers to give them answers in advance, and other things that are spiritually no different. ChatGPT just exposes the problem.

If you're going to heavily incentivize credentialization through all layers of our culture, you're going to get students that only care about getting the certificate, not learning.

I wrote a cheat sheet version of my Using ChatGPT is not bad for the environment post. It's paired down and more focused on simple responses to common objections, without a long intro or my background environmental philosophy. andymasley.substack.com/p/a-cheat-sh...

"They teach us to call evil ‘evil,’ a simple change but one which leads us to choose evil less." From @adapalmer.bsky.social's 'Perhaps the Stars'. The line has stuck with me since I read it and I've tried to live that practice more.

This is the essence of one of my critiques of the AI doomer argument. Being able to think super hard about a problem is a necessary --but not sufficient -- condition for solving new problems. You still have to experiment along the way to map out the unknown unknowns.

Every time I see a take like this I'm reminded of the librarian at my high school trying to rip a computer keyboard out of my hands when she saw me on Wikipedia in 2004.

This is the root of many things that I find distasteful in our culture. Cruises, Disneyland, Theme Parks, expensive kids' birthday parties. They're all these very refined, pre-packaged experiences. There's no room in them for individuality and exploration.

Free trade, free movement, vaccines work.

Any time I hear someone pine for a post-scarcity economy I think about how I can buy 41,000 calories of flour for less than $10.

Promptus Miserabilis

I’ve found LLM chatbots similarly helpful for topics where I have low context and don’t even know what terms to look up. It’s a really helpful first pass for picking up enough terms and context to ask more questions or do a search.

Mundane uses of LLMs My son had an appointment with an allergist yesterday to check his sensitivity to certain allergens, and see if anything has changed in the last year.

Like you, I occasionally experience overstimulation. Usually it comes from my small children, who overwhelm me with their noise, presence, and demand for attention.

And I'm depressed to see how many people are eager, even joyful, at shutting down the border to the neediest people.

Like you, I have struggled with making Cacio e Pepe, that glorious Italian Mac & Cheese variant. My cheese would always clump and the final sauce would be gluey and chunky instead of smooth and unctuous.

Another mundane use of LLMs: providing context to posts you don't understand.

@zachweinersmith.bsky.social did you see this? Re: philosophy of AI media.ccc.de/v/38c3-self-...

Einstein still has to live somewhere, and Einstein with access to daycare and an assistant and a skilled plumber is way more productive of Einsteiny stuff! By opening up immigration, we let people's needs decide who comes, rather than using arbitrary quotas.

I think a lot of people don't realize how wildly useful LLMs like ChatGPT are in everyday life, so I'm going to document the mundane ways that I use them.

In related news, I just looked up numbers for the cost of menstruation and found the staggering amount of 17000€/lifetime…which includes 5€ of painkillers, 9€ in underpants, 10€ in sweets and 8€ for misc stuff like hot water bottles per month.

Oh my stars, I have been tearing my hair out on this exact problem. There are charities that try to address "diaper need" for families, and they often cite a figure that diapers for a single child cost $1200 a year. www.pdxdiaperbank.org